The ABINIT Users Mailing List ( CLOSED )

[Damien Caliste <damien.caliste@cea.fr>]

Hello,

Le 02/04/2009, cszakacs@mun.ca a écrit :
> Or the acell parameters are to high?
I would suggest to give a look to the BigDFT paper
(http://link.aip.org/link/?JCP/129/014109) where you have a convergence
plot with respect to the box size for a cinchonidine molecule. You will
see that the plane wave basis set is not always a good solution.

Since ABINIT 5.5 (or .4) you can you the formalism described by the
above paper inside ABINIT. This use the usewvl keyword.

If you're interested in structural optimisation of clusters, I would
strongly suggest to use the wavelet formalism instead of the plane wave
one. It is quicker (see the table beside the plot in the above paper)
and less memory consuming.

For the use of usewvl keyword and related wvl_* keywords, don't
hesitate to post on this list.

Damien.